SEO by the SEA feature some interesting research and an article on Google’s most popular and least popular top level domains. William Slawski went and checked the amounts of results that Google stored for each TLD and created a nice table of the results. Quite interesting. A great idea.

Now who’s got the same data for searches?

[added Jan 14th]

I noticed something interesting about the data – it seems .CA ranks higher than .DE, which seems a bit weird, considering that there are way more .DE sites.

[via SeachEngineJournal]

Peter just pointed me to an article adding to the speculation about Google becoming a registrar: Jim Boykin’s Internet Marketing Blog » Blog Archive » What does Google know about your domain names? by Nick Wilsdon

Nick points out that Google does gain additional privileges by becoming a registrar, such as carry(ing) out high volume automated queries.

It would allow them to track the age of domain names and connect likely networks and portfolios. They would see when a domain name changed ownership and allow them to adjust the value of that domain in their engine.

While their accreditation status would probably giving higher query limits for the registry whois, this whois only contains limited information. For the biggest gTLDs, COM/NET, the whois only contains:

• the domain name
• the creation date
• the expiry/ date
• the last updated date
• the nameservers
• the sponsoring registrar
• the sponsoring registrar’s whois server

The most valuable pieces of information for Google would be the nameservers, the creation and expiry date. In the past there has been some speculation (based on a patent google filed) that they give higher evaluations based on the length of the registration term. I personally do not thing that this is such a great idea, considering that a lot of hobby websites are only registered and renewed year by year. Plus it looks like some names are now actually being sold to new owners without being re-gistered, so they are actually never dropped.

The registry API (at least for the gTLDs) will not give Google access to any additional information – as long as the domains are not registered through them.

I’d like to clarify something: In order to track actual owner information for domains (at least for COM/NET), they also have to query the whois-servers of the registrars, who in turn have their own rate limiting. Don’t get me wrong, it is a common request between registrars to increase the whois query limit, but in most cases this request is only granted in order to facilitate transfers from one registrar to another (for transfers you need a confirmation from the current admin contact listed on the domain). Since Google is not actively selling domain registrations, I am not quite sure how many registrars would increase their whois query limit for Google.
Also it is important to note that getting accredited for ccTLDs is a different process from being accredited for gTLDs, mainly because most countries create their own rules and API systems/access methods.
As Peter also reminded me, registrars also have access to the list of names that will be dropped for the registry, which would indicate a change of ownership.

Danny Sullivan wrote up a summary on the current ‘ads on parked domains’ discussions, which were started off by ‘research’ (MS Research: Typo-Squatters Are Gaming Google) published by Microsoft (I am still wondering if this means that they are not going to try to get into this market with their Adcenter offering.).

I was planning on a summary, but his article give me a good starting point – while I don’t necessarily share his opinion:

Search Engines Making Millions Off Type-In Traffic From Domains

It’s not that domain traffic is necessarily bad nor that domainers are necessarily doing anything wrong. But there’s a difference between the person who did a keyword search as opposed to someone who did a navigational guess (a type in) versus someone who sees ads placed contextually on pages. Advertisers deserve the transparency of doing these as separate buys.

For a while I thought about this the same way, but is there really a difference? A lot of users seem to use domain names as a replacement for search queries. A generic domain that is typed in by the user would probably be the most likely to convert. Someone else follows a link that points to a domain that has been grabbed in a domain drop, will it convert? If they type a domain name of a brand wrong, they might end up on a PPC page. Will they find the brand name they were looking for in an ad on that page?

It would be great if someone (maybe one of the aggregators?) could provide data that shows that domains do work well for the advertiser. I am convinced that there are numerous examples out there. I would just love to see some research – it could teach us a lot. Maybe it even makes sense to split the traffic, based on it’s origin. And of course we also need to make improvements on the other end, after the click on the ad. Bad conversion rates can also be related to a bad and non-targeted landing page.

For the advertisers this is really about the quality of traffic, or even more specific, the likelihood of conversion. Which mindset is the end user in? Which part of the sales cycle is he in? There are many more questions that need to be answered so we can ensure that all the parties are treated fairly in the process. How and where should conversions be tracked, and how much higher would the payout be for a click that converts? What if the conversion takes place 6 months after the click?

Even if a space is very non-public up to a certain point, it’s always a good idea to try to steer how and where this topic will appear publicly, especially if you have a suspicion that there is a strong chance for a negative twist. Now it’s now important to control the damage, which can be a long and painful process.

Some more posts/discussions about this topic:

http://www.searchenginejournal.com/index.php?p=2694
http://www.circleid.com/posts/whats_wrong_with_domain_names/

In the end, I believe that the short term outcome might be that advertisers will be able to pick between content syndication, search and domain advertising. This will have a negative effect for a lot of domainers, at least in the short term. Long term it might lead to a tiered structure where the payout depends on different factors (a combination of of pay per unique, lead, action, click etc).

Happy New Year. Since it’s always a good question to ask at the beginning of a new year: What will this year bring for domainers?

Here are some things that come to mind:

• publicity (this will include more negative coverage)
• newbies
• new investors
• different levels of domain/site development
• new ad models
• portfolio sales and partnerships
• launch of .EU (more on this later)

What are your expectations/plans for the new year?

Types of Domain Traffic (Search Engine, Direct Navigation, Expired Links)

This is an attempt to summarize the different origins of ‘natural’ domain traffic on the Internet. I am concentrating on ‘natural’ traffic here, so I am not including PPC ads or similar means to generate traffic. I came across three measurable main types of traffic that can be generated by a domain, but I think that there will be more categories added as tracking improves – please feel free to comment, I’d love to hear from you.

1. Search Engine Traffic

This is traffic that originally is generated by means of a search term on a search engine. The user is then referred to the target site from the search results.

You can identify this type of traffic by taking a closer look at who refers your users to your website. In the logfiles you will also be able to identify the queries which lead the user to your site. Often this can be used in order to optimize your site so your users find the information they are looking for as quickly as possible.

In order to improve this type of traffic, one often needs to builds an ‘authority site’ on the topic that goes along with the domain name. This can be done by either blogging, re-posting excerpts from other sites (possibly in conjunction with comments etc.). Another possible source for content are free article websites. To improve this type of traffic, SEO (search engine optimization) strategies can be employed.

2. Direct Navigation or Type-In traffic

The user types the name directly into the address bar of the browser, accesses it from a bookmark or has it set as the browser’s homepage. Many users use domain names as a search term (I know that this is something people often don’t believe, but the right name will generate a significant amount of qualified type-in traffic). So a user actually looking for used cars in Toronto, would directly type “torontousedcars.com” into his browser’s navigation bar. Those users often do not distinguish between a search and entering a domain name.

It would be very interesting to differentiate between bookmarked/homepage and actual typed in traffic, but this is difficult to track without using cookies. For example you would be able to see if an Internet Explorer user bookmarks your site (the web browser will access the ‘favicon.ico’ at this time), set a cookie and identify further visits from the same user by querying the cookie.

Another possible origin for this type of traffic includes Instant Messaging and mail, which could form an additional “Word of Mouth” Category. As Brian (my co-worker) pointed out a couple of days ago, this traffic could also originate from print, TV, flyers and other forms of advertising (I see the URL, I remember it or write it down and type it in at a later time). Currently the only way to track this type of traffic separately would be to use a unique URL/domain for every publication.

3. Typo Traffic

This category monetizes miss-spellings of popular high-traffic domains. If investing into this type of domains, one should be very careful to avoid trademark conflicts etc.

4. Linked Traffic (expired/non-expired)

Domainers will pick up domains that have previously been used for an active website. Often these sites experience residue web-traffic that originates from previous links to the site and also from search engine listings. When the original registrant fails to renew their domain name, the new owner buys these names at the dropped names services. Good names will mostly be sold in an auction.

You will be able to identify this type of traffic by checking the referrer URL in your log files. If a site with this type of traffic just serves PPC ads and no content, you will often see the traffic drop off after a couple of months (as search engines might drop them from their index). If this type of site is developed, it might be able to maintain the benefit from the previously existent site.

Blogster.com Purchased for $100,000, Making it a Top Domain Name Acquisition in 2005

While we normally do not report every posted Domain Transaction, I think it is very interesting to note in this context that more and more of the domain name sales nowadays are accompanied by a Press Release. The public mentioning of those Sales helps to bring more visibility to our industry and can possibly help drive the prices up (which of course is not necessarily in the interest of the people still building their portfolios).

Due to the recent confusion, I would like to add a link to the Wikipedia Definition of Cybersquatter/Cybersquatting:

Cybersquatting is a derogatory term used to describe the practice of registering and claiming rights over Internet domain names that are, arguably, not for the taking. The Cybersquatter then offers the domain to the person or company who owns a trademark contained within the name at an inflated price, an act which some deem to be extortion.

The important part in this definition is that a Cybersquatter knowingly infringes on the rights of others for financial gain. While those people still exist today, the large portfolio holders have gotten wiser.

Most of them go for generic terms, they shy away from anything that could be considered a copyright or trademark infringement. Of course they might accidentally still pick up the occasional trademark (after all they cannot screen all of the domains they buy for trademarks in all the countries of the world), but most of them are also quick to give up the domain should they receive a Cease & Desist letter or another complaint.

By the way, instead of always sending Cease & Desist letters right away, it might be well worth your while to contact the current domain owner and see if there is a more peaceful solution.

(picture from sxc.hu)

In Domain Editorial (BETA) » Another Alternate Root? I asked:

Will Unified Root’s Nameservers be storing any queries for currently non-existing domains? If so, what will be done with that data?

I received the following answer from their press contact a couple of days ago:

(some typos were corrected)

I have asked UnifiedRoot for response. They say: Root servers do not store any query history or data. Data collected on root servers are purely performance data and does not contain any information valuable for sales and marketing.

While we obviously do know that those queries do contain a possible value for determining typos and other available traffic domains, it seems that this new alternate root does not (at least not publicly) know anything about these potentials. Maybe they should read this blog

As noticed by Sheperd (thanks!), NetIdentity/Mailbank has added Icon categories and a search to their lander pages. If you are not familiar with their business model, they own many family names as domains and some generics and offer email and other services based on those domains. The services include email, hosting, blogging and additional domains.

For many years a lot of people in the domain/PPC industry have been saying that NetIdentity should PPC park at least some of their generic domains. I also noticed that some of their domains are actually now parked with DomainSponsor and just have a small header on top pointing to their product offerings:

Some domainsponsor examples:
http://zoologist.com
http://snowboarders.com
(this domainsponsor design always reminds me of the Digimedia sites)
http://deepsnow.com

A “normal” new landing page:
http://fick.com (ever considered putting German PPC ads on this domain )